Medical treatment for metastatic melanoma (MM) has primarily focused on biological therapies designed to mobilize immune, effector cells that recognize and destroy cancer. Treatment with interleukin-2 (IL-2) has shown dramatic clinical efficacy in a minority of MM patients. The absence of benefit in the majority of patients may be due to tumor resistance and/or inadequate effector cell response. Adoptive cellular therapies with lymphokine activated killer cells (LAK) or tumor infiltrating lymphocytes (TIL) and IL-2 have been employed to overcome inadequate effector cell response. Unfortunately, randomized trials in MM patients have failed to confirm benefit for the addition of cellular therapy to high dose IL-2. The failure of adoptive cellular therapy may be related to persistent tolerogenic factors or inadequate space for cytotoxic effector cell proliferation in the lymphoid compartment. Chemotherapy induced lymphodepletion followed by high dose IL-2 with adoptive transfer of selected ex vivo expanded TILs has induced significant responses in approximately 50% of MM patients. Laboratory studies have demonstrated that lymphodepletion alone creates an anti-tumor immune environment capable of inducing regression of tumors. Following lymphodepletion, homeostatic repopulation of T-cells occurs with antigen-specific T memory cells and naive T-cells skewed to antigen-specific effector cells by antigen presenting cells. Granulocyte-macrophage colony stimulating factor (GM-CSF) enhances number and function of antigen presenting cells. Thus, we hypothesize that lymphodepletion with cyclophosphamide (C) and fludarabine (F) will provide a fertile environment for de novo regeneration of antigen presenting cells by GM-CSF and IL-2 driven melanoma-directed cytolytic lymphocyte populations with subsequent clinical benefit. We propose to test this hypothesis in a 2-stage phase II study. Eligible and consented patients with MM will undergo treatment with high dose C (day 1,2) and F (day 3-7) followed by 2 5-day courses of high-dose IL-2 separated by 9 days rest. GM-CSF will be given starting on day 8. In responding patients, a second course of IL-2 will be administered. We propose to determine 1) the complete and partial response rate to treatment, 2) the toxicity profile of administering this regimen, 3) the kinetics, phenotype and function of lymphocyte and dendritic cell recovery, 4) time to progression and survival and 5) the relationship of lymphocyte recovery and objective clinical responses. [unreadable] [unreadable] [unreadable]